Infective endocarditis is a lethal disease that is poorly understood. Despite one-year mortality rates of 40% there have been no randomized trials examining treatment strategies to reduce endocarditis-related complications and death. The primary objective of the proposed study is to facilitate the development of an evidenced-based approach to manage patients with endocarditis. Specifically, we aim to: 1) characterize in detail, contemporary patients with endocarditis and examine regional variation in their characteristics; 2) identify predictors of thromboembolic events and death; and 3) identify potential host determinants of complications in endocarditis As part of the study, we will develop prognostic models using data from the Duke Endocarditis Prospective Cohort Study, the longest running study of its type. These models will be validated in an independent patient population enrolled in the International Collaboration on Endocarditis (ICE), the first study of its kind. The findings of this study have implications not only for clinical care but will also provide a foundation of knowledge and network of investigators from which subsequent randomized trials can be conducted.